Login / Register
 
Building The Most Advanced DAW On The Planet
Closed
Subscribe
Antonis2007
Thread Starter
#61
26th December 2012
Old 26th December 2012
  #61
Gear maniac
 
Antonis2007's Avatar
 
Joined: Feb 2007
Location: Mountain View

Thread Starter
Antonis2007 is offline
Merry Christmas and a Happy 2013 !
Let 2013 change music for good !
#62
26th December 2012
Old 26th December 2012
  #62
Gear maniac
 
gmx750's Avatar
 
Joined: Nov 2011
Location: Azores
Posts: 198

gmx750 is offline
Are you talking about hardware and sotfware sollution or optimised software running on existing hardware?
The latter makes sense, and If that's what you are aiming for, I would suggest stable unix based kernel. striped down to essential services. Optimised audio APIs.

sketch up the baseline for the core OS, main goals for audio and I/O performance (system buses and internal / external interfaces).

gather up community feedback and oppinion and start sketching your software.
focus on the simplicity of picking up an acoustic guitar and letting your creativity do the rest... software and human interfaces should allow for the same.

my 2 cents,
GMX750
#63
26th December 2012
Old 26th December 2012
  #63
Gear nut
 
FYUnited's Avatar
 
Joined: Apr 2011
Location: Kotka, Finland
Posts: 126

FYUnited is offline
I'm getting tired of software GUI's trying to look like hardware. Like Undertone and a few others have pointed out this is utterly redundant (unless running hardware through the software maybe) especially when doing things completely in the box.

I'd like to be able to get rid of tracks altogether. Fluid, easily assignable audio elements combined as I see fit. Add to this more meaningful info on the content of these elements (pitch, volume color coding and more) to help you work. FabFilter is going this way, but only as plugins. I want a DAW that does these things. Kind of like Logic Environment but without the steep learning curve.
__________________
Aaro, FY United
A hobbyist: OSX 10.7.x - iMac i7 3,4 GHz - RME Fireface 800 - Logic 9 - FabFilter & ValhallaDSP - Genelec 8020s - Misc gear & too many guitars
#64
26th December 2012
Old 26th December 2012
  #64
Lives for gear
 
shponglefan's Avatar
 
Joined: May 2010
Location: Great White North
Posts: 2,486

shponglefan is offline
Quote:
Originally Posted by Antonis2007 View Post
Do not worry about h/w and s/w capabilities. The beauty of it is that it can work with 'anything' - we have to keep it Open !! I believe that it is open by it's nature and that no-one will try to 'block' evolution. I hope that humanity and industry moved forward from 'buying' software companies in order to 'create exclusivity' inside a platform - proved by Sad facts that this approach do not have a sustainable future. It created enormous pressure to the people who try it and the market. and the outcome might be few so called 'best products' but in the long run it moves things backwards !
We need a sustainable future.. not on papers but in reality ! I believe we can have it !
I tried reading this paragraph three times, and I still don't know what it means...
#65
26th December 2012
Old 26th December 2012
  #65
Lives for gear
 
Joined: Jul 2005
Posts: 3,038

Bassmankr is online now
This all seems pointless to me as the tools to do all this already exist and are affordable. Stop painting yourself into a corner thinking it all has to be all digital and interconnection has to be through fast ethernet. Just use a cluster of standard DAW PC's (around $1k each). Assign each PC to a certain number of instruments / parts of the orchestra. Send midi from your keyboard controller to a midi patchbay with outputs to your main computer and each cluster computer. Treat each of the cluster PC's as a dumb sound module (the same way you would treat a stand alone synth with no musician playing it or a rack synth. You only need to send it triggering info to play it's midi part). Run the audio outputs from each computer to an analog mixer. With just a standard PC network at standard speed you can use remote desktop software to start up each cluster DAW and select it's instruments from the main PC without needing additional moniters but you can still get fancy there if you wish with a good KVM switch.

This method has been around a very long time and works fine. With each new generation of CPU, the expansion of vasts amounts of memory now available for samples, and thanks to the new 64 bit operating systems you just need fewer PC's in the cluster to make it happen today.

With all your instruments setup on PC's and outputing to the analog board for monitering and mixing there is no break in the "creative" part of making music as every instrument is available. By the way you can accomplish the same thing with real instruments all going to the analog mixer too but you need real musicians then (that method has been around even longer).

No need to re-invent the wheel despite Grant money or school paper to write.
#66
27th December 2012
Old 27th December 2012
  #66
Lives for gear
 
Joined: Feb 2005
Location: on the couch
Posts: 1,655

MarkusColeman is offline
Quote:
Originally Posted by Bassmankr View Post
No need to re-invent the wheel despite Grant money or school paper to write.
I disagree with this statement, but not the essence of both your posts.
Reinventing the wheel is the most important part of this technological revolution, but it's not enough to "stand on the shoulders of giants" (including the OP) when your goal is to fly.

Here are some comments of mine:

1.) HPC is equal to "High Performance Computing". In layman terms this means, that you have a computer, which isn't wasting resources on unnecessary services and which utilizes the best technology to do so. The newest algorithms ran on a 30 year old Apple II can outperform today's world's fastest super-computer. Ponder this for a while. In fact, that's all you have to really think about, when it comes to "HPC".

2.) You claim to have access to the world's fastest networking technology. I doubt that, because I would know if that were the case. Laser-based connections between "CPU-layers" are just in their infancy, so I doubt you have years of expertise working on implementations.
I guess we are back to network cards, copper and wireless then, huh?

3.) Ask yourself this very tricky question: is there a problem that requires parallel computing and can't be solved by simply utilizing faster sequential computing? Even parallel sorting algorithms run in steps.
If you can show me such a problem, you will be the first.

HPC computing is so popular because of the availability of cheap hardware AND because we've realized that we can make cheaper cores, but we can't make the cores run faster without paying much more for it and hitting the ceiling. We are having 48-core systems now, because we can run them at 2-3GHz and can not run a single core at 500GHz without it turning into a Black Hole. of

4.) Why would I need networking when I can use RAM-based drives NOW and for cheap? If I buy 10 of the fastest SSDs on the market, buy a 12 port PCI card, I will probably hit the area of 6GB of data-read per second. That's roughly THIRTY SEVEN THOUSAND TRACKS OF AUDIO at 16/44.1 kHz.
Again, that's roughly 37,000 tracks of audio at 16/44.1 kHz. SIMULTANEOUSLY.

That's too expensive for you for 30+ THOUSANDS of tracks? All the SSDs and the card, etc...? Buy yourself 2SSDs and 64GB of RAM and create a 32GB RAM-drive, which with new RAM modules will allow you to roughly meet the 6GB/s of data-read above. For $750. On a standard mainboard, using a standard CPU, standard RAM modules, standard HDs.

Even at 32bit/384kHz, you will have enough of sound sources at your fingertips.

What was that proverb again about blind lumberjacks and not seeing woods for trees?

5.) You ask yourself, if you understand music and sound. But the question is: does your guitar understand what flamenco music is? No? Then ponder this for a while. Said guitar is made to sound like a flamenco guitar, or to be played like a flamenco guitar or the sample library utilizes some clever layering so that you are able to play it on your keyboard like a flamenco guitar. Or you are simply recording a flamenco guitarist played by a guitarist who knows what flamenco sounds like. You see how complicated such a simple undertaking is? And that's the easy part, trust me.

The hard one is the following: does your DAW understand what it does?
Does it understand MUSIC?

The answer is clearly NO. Solve that problem and stop talking about reinventing the wheel, while having a rocket in mind and flying to the Moon.
And that's what you need to do, instead of working on "alternatives to a circular wheel design".

6.) All of that is comparable to Newton's undertaking of creating the mathematics necessary to "harmonize" the orbits of the planets back in the 17th century. He reached his limits (because lets face it, even Newton (unlike Tesla) wasn't ahead of his time) and then started to talk about God and Intelligent Design, because he couldn't imagine how something so complicated could be solved. As we know today, Archimedes already developed a lot of the math Newton ended up re-creating almost 2,000 years later. But Newton couldn't have known that, because Archimedes' works were rediscovered in the 20th century. Re-inventing the wheel again (this time, not on purpose).

What I mean by this? Two things: for one, look into the past. There are already designs based around creating music, where the software understands the theory of what is played and what will be played in the future. Be it scales, rhytms or effects.
Number two: I used the example of Newton, because the problems he was dealing with and him claiming it was unsolvable and "God's work" ended up being solved easily just a couple of decades later. So why did Newton talk this nonsense? Because he couldn't believe, that he hit the ceiling of his abilities. So don't mistake your limits for the limits of the world.

And realize, that you are basically trying to create a Deus ex Machina.

Speaking of "celestial harmony", a field once taught at a single university at the beginning of the 20th century. "CELE(stial har)MONY" ended up being the company of a certain German music enthusiast, which brought us Melodyne. And lately, DNA, or the algorithm to be able to reconstruct music from its basic building blocks (hence the convenient abbreviation). "Reductionist re-synthesis" could be the name of the field you have to master next, if it existed. Maybe the man in question at Celemony will help? Who knows?!

7.) Lucky number 7. The Reason for Live is Logic. Sorry, couldn't stop myself. Seriously though: music makes us happy. Making music makes us happy. For different reasons, given different scenarios.

Sometimes you sit there, alone, your guitar in your hands or a keyboard or a set of drums. And you start to make music. You know what the problem is? No matter the software, you will always be a one-man-band. If you want to revolutionize making music, then you have to change the process.
And to change it, your DAW must understand music. And it must be able to ACCOMPANY you, while you play. You can be the guitarist, the vocalist, the keyboardist, the drummer.........or the conductor. But where is THE BAND that can play along, when the only person in front of the computer is YOU?
That's one part of the problem and not really something you will be able to solve BY YOURSELF.

The other problem (and the one which should be solved first) is; why can't my DAW understand my REAL BAND? Why doesn't a DAW exist which lets me and my three other friends grab their instruments AND/OR tablets and choose their instruments and we jam together, while the "main DAW" communicates, records, manipulates and creates the audio output?
A DAW should be able to do that. Let PEOPLE (read: plural) make MUSIC (read: understand what's going on and create the sound sources for it) TOGETHER (read: simultaneously, while conducting the session).

That's it for now.

As you can clearly see, the problems are not really hardware-based. Or the problems that exist can be easily solved using the right hardware and the right approaches.

The challenges to REALLY give musicians the freedom to create music the way they imagine it, are almost exclusively in the "reinventing the wheel" arena of ideas and software.
Antonis2007
Thread Starter
#67
27th December 2012
Old 27th December 2012
  #67
Gear maniac
 
Antonis2007's Avatar
 
Joined: Feb 2007
Location: Mountain View

Thread Starter
Antonis2007 is offline
Quote:
Originally Posted by Bassmankr View Post
This all seems pointless to me as the tools to do all this already exist and are affordable. Stop painting yourself into a corner thinking it all has to be all digital and interconnection has to be through fast ethernet. Just use a cluster of standard DAW PC's (around $1k each). Assign each PC to a certain number of instruments / parts of the orchestra. Send midi from your keyboard controller to a midi patchbay with outputs to your main computer and each cluster computer. Treat each of the cluster PC's as a dumb sound module (the same way you would treat a stand alone synth with no musician playing it or a rack synth. You only need to send it triggering info to play it's midi part). Run the audio outputs from each computer to an analog mixer. With just a standard PC network at standard speed you can use remote desktop software to start up each cluster DAW and select it's instruments from the main PC without needing additional moniters but you can still get fancy there if you wish with a good KVM switch.

This method has been around a very long time and works fine. With each new generation of CPU, the expansion of vasts amounts of memory now available for samples, and thanks to the new 64 bit operating systems you just need fewer PC's in the cluster to make it happen today.

With all your instruments setup on PC's and outputing to the analog board for monitering and mixing there is no break in the "creative" part of making music as every instrument is available. By the way you can accomplish the same thing with real instruments all going to the analog mixer too but you need real musicians then (that method has been around even longer).

No need to re-invent the wheel despite Grant money or school paper to write.
Before iPhone - you could find touchscreens on communication devices (30 years ago) but that does not mean that iPhone did not change everything.

If you are looking for 'Grand Money', fancy 'paper' and marketing hype, look at different places.

HPC and Music is 'a combination' of the Brainpower of the most innovative people on the planet. People who make software and hardware masterpieces for the last 5-10 years. But they do not make it for HPC and music (HPCmusic is how I like to call it). They use a different mindset to design their tools. They think 'in the box'. They think by using the personal computer principles. They use their tools as a side-car to sell other tools they have.

I take all those innovations they make and I just combine them because I have a certain vision of what I want to achieve. I am working on creating a market that does not exist. A solution that will move things forward. And not just writing some code on C++ nor i am just connecting computers together with KVMs, custom ethernet or midi - I think many people do that already.

You can be sure that in order to shape that vision - i followed all the suggested and more importantly the non suggested ways to use computers and networks. I pissed of a lot of people.
I made music, using all of those technologies. You can not envision the future by making mathematical calculations nor by writing some clever code.
You need to dive in to reality. you need to understand psychology, you need to understand music (when you are creating for music professionals) and sometimes... and this is an example of our lifetime ... you even need to understand 'typography'. You need to have taste and culture. And always be ready to fail - you must respect failure as much as success (and maybe more). That makes you humble and peaceful. It's OK - I might be wrong

Hopefully, i will inspire someone else - our ideas will have sex and the world will become a better place ! That's the point !

It is that strive with reality that 'gave' me a solution. It comes to you because you dare to question.

And I am here to share it.

When the time will come and i will share it, it would be up to you to try it and then decide if it works and if it can move us forward.

It will either create a new market and at the same time a new future for music or it won't.
It is that simple.
#68
27th December 2012
Old 27th December 2012
  #68
Gear nut
 
Joined: Nov 2012
Posts: 96

ArKay99 is offline
Thinking outside the box...
Since we are dreaming, why not keep the analog signals analog instead of digitizing them? Store them in analog format and you won't need ADC's and DAC's.

Back in the 70's I wrote a 32 track sequencer for the Atari 1040ST. the 'JackIntosh'. I had the OS code and assembly instructions for all the drivers. It was an eye-opener to see all the overhead to be borne by those drivers in trying to maintain some sort of abstraction. I was able to re-write the midi input and output routines with much less than half the instructions and keep them 100% reliable. Today, Windows has all sorts of abstractions between the GUI and the hardware. It takes up a lot of bandwidth. I'm sure the Mac has similar constraints. In addition to wasted bandwidth, digitizing systems suffer from latency. Even a 1 sample latency will cause comb filtering throughout the audio spectrum when summed with the input signal.

In the 80's I worked for AMS and we used Transputers. Created to eliminate the Von Neuman bottleneck... an eq would have a few transputers and one would calculate the coefficients in real time and feed those off to another doing the multiplication of the signal and those coefficients and that was all fed to a transputer doing the routing and a fourth was handling the interfacing. They all communicated on a 20mhz serial link. You could have transputer farms and configure them to do anything you wanted. However the main problems we dealt with was I/O latency and converter resolution...and jitter.

So, the 'ideal' DAW would have 2 main areas of implementation. On the hardware side, you'd have to start with zero latency, zero digitization artifacts, zero jitter, infinite scalabilty, and the ability to interface with every other piece of 'electronica', be it audio, video, or other, and be able to lock to any time consideration.

On the software side it should be able to easily manage the hardware layer through abstraction. It should also be able to be easily configurable to any mode of operation. Perhaps a configuration can be built by dragging icons as representations of DAW objects from a palette to a canvas. Even those objects could be built out of 'draggable' components. Plugins would be a thing of the past as 'characteristics' could be dragged onto a canvas for the effect.

Then the whole thing should be 'scriptable'. Give the owner a way to drag and drop objects together to create new 'modules' of functionality. The old Next platform comes to mind here. Say we need a custom serial interface. You should be able to 'construct' it by dragging icons onto a canvas that gives the serial interface it's properties.

Why stop there? Say I want to construct a string section, or the ultimate drum kit, played by any drummer I have a 'characteristic of. I could make a kit of any number and type of component by dragging to my canvas its copmonents, then the room it's played in, and the console or no console it is to be played through, and the 'fur de lance', I would grab say, Steve Gadd or Simon Phillips and drag his playing style onto the canvas. I wonder what Jeff Beck's style would do to a drum set? Of course, those styles would be able to be modified to.

Then we'd have the actual recording, playback, and mixing. As suggested before, why have the constraint of having to use a console as an abstraction. Since we can define the GUI to be whatever we want, why not perceive the whole experience as a 'performance'. Instead of a console you could have, say, a stage, with the parts (performances) dragged to the stage. Put in positions on the stage. Left, right, forward or back. The drums on risers...the bass amp in the bass drum? Put your listening position in the upper tier, or wherever is 'desired'. Or maybe the venue is your favorite club, a parking garage, your friend's basement, or the Sydney Opera House...

Most of the above is achievable with today's technology. I believe to make it happen would be a dedicated OS that is Open Source, but focused on audio pathways and associated processes being as lean and efficient as possible. It should be language friendly. i.e. written in C++, Basic, or whatever. It should be compiled for runtime efficiency. Some company would need to make a motherboard that is optimized for this platform with lots of PCIe slots, Thunderbolt ports and gigabit ports. There will need to be plug in cards with distributed processing all over them. The hardware should be addressable by the OS and its scalabilty should be transparent to the OS. If it's modularized, the OS should know what hardware components are present and only include the software layers required to run the hardware. This can be done by a compile at upgrade or install. The components should be distributed and configurable with an abstraction layer. The media, at least with today's technology, can't be stored analog unless we go back to tape, but current digitizing technology is pretty good...should I say, good enough? Media needs to be stored in as safe and retrievable manner as possible. Redundant Raid?

Lastly, DAW's are not only for recording music, so a drag and drop interface to construct whatever 'surface' is desirable for the task at hand could actually be accomplished. Next was doing it years ago. Then visual monitoring should be configurable for however many monitors are required.

This is just a start, but I believe the key to the ultimate DAW is configurable abstraction at every layer. As seen up to this point it's been impossible to achieve. I believe that is in part due to the attempt to utilize non-DAW-dedicated OS's.
#69
27th December 2012
Old 27th December 2012
  #69
Gear addict
 
Tha Govna's Avatar
 
Joined: Mar 2009
Location: Jacksonville, FL/Rotterdam, Netherlands
Posts: 318

Tha Govna is offline
Quote:
Originally Posted by Antonis2007 View Post
What would you want from it?

Time to dream bigger.
I'd want it to look and perform like Protools, only better and 64-frakin'-bit finally.

That's it.
__________________
"I don't have to dodge shots, I'm the shooter".

Larry Otis.
#70
27th December 2012
Old 27th December 2012
  #70
Gear Head
 
Joined: May 2009
Posts: 59

daexpert is offline
Quote:
Originally Posted by ArKay99 View Post
On the software side it should be able to easily manage the hardware layer through abstraction. It should also be able to be easily configurable to any mode of operation. Perhaps a configuration can be built by dragging icons as representations of DAW objects from a palette to a canvas. Even those objects could be built out of 'draggable' components. Plugins would be a thing of the past as 'characteristics' could be dragged onto a canvas for the effect.

Then the whole thing should be 'scriptable'. Give the owner a way to drag and drop objects together to create new 'modules' of functionality. The old Next platform comes to mind here. Say we need a custom serial interface. You should be able to 'construct' it by dragging icons onto a canvas that gives the serial interface it's properties.
I think you're looking for a modular environment such as reaktor, max, or tassman.

I don't think that's what most people want though.
Antonis2007
Thread Starter
#71
28th December 2012
Old 28th December 2012
  #71
Gear maniac
 
Antonis2007's Avatar
 
Joined: Feb 2007
Location: Mountain View

Thread Starter
Antonis2007 is offline
Some of you may know that I have an awesome friend - Trey Ratcliff.

Trey got a picture of Hans Zimmer's studio back in January 2010. I upload a 'crop' of the photo (sorry Trey) and so you can see the most advanced DAW (Digital Audio Workstation) on the planet, until that date.

This is a pure over-engineered hybrid monster. Dark Knight Trilogy, Inception and Pirates of the Caribbean were created inside this system. Hans Zimmer has a huge collection of analogue instruments - so he based his workflow around them. You can count around 15 nodes on a 1GbE network clocked at 44.1kHz. You can see that it needs clocking from a dedicated device...etc.

If I was on tight deadlines for a blockbuster movie and I had the money to buy anything I wanted... this is what i would do back in 2008. But I am not the best composer in the world... so from 2007 (when i build my studio) I had the time to explore different setups without being on very tight deadlines. Around early 2009 I politely pushed away all of my clients from the studio. I deliberately became isolated from my surrounding (i did not thought it would be that hard). I could not afford to go out of the studio. I had enough money only for food and electricity bills - nothing else (for almost 4 years). I had some rare visits from friends - i could not have a personal life.

Jogging in the forest and talking with my cyber-friends was my only escape. But jogging was not enjoyable either... i felt like i was trying to run away from reality. I was honored by the fact that my cyber-friends were awesome people doing extraordinary things (this gave me such a boost).
Every little email I had... every little 'word of hope' and encouragement would kept me going.

I focused on my work with Trey Ratcliff. I could explore soundscapes that try to match/recreate the depth of HDR photography into the sound. I did some crazy things - in some point i was mixing a '300 tracks - 3 minute' song for 4 months (and for that same song I did all the arrangement/re-composition and sang on it). 'Depth' and 'Dynamic Range' became my roommates. I knew what I wanted to hear and i was looking for it like a madman. I failed so many times. Every time i was creating a sound that felt right on top of an HDR shot - i knew i was in the right direction. When 'every in the box approach failed to give me what i wanted ... i broke the rule - but not by added analogue tools and making it a hybrid system. I added other boxes. I was not in the box any more.

And that I came with the profound idea that out of the box, does not necessarily mean analogue/hybrid system. Out of the box is about analogue qualities and power. It made me look at my $50k Mac Pro with HD6 and 4 UAD cards like it was a piece of technology 'confusion'. And from my perspective - it was. Not that DSP is not a nice technology - it is and can be an added value to any system... but... the BOX was/is the problem. And i was looking at a viable solution for my 'out of the box' dreams.

I started asking myself questions like:

- what a plugin could do when it is not designed to work with 'in the box' limitations?

- which are the best interconnection methodologies when you have 'many boxes'?

- how can I take supercomputing technologies and bring them down to the music production field?

- what HPC can do for music?


I tried to let you know what I was thinking back in 2010.

We are almost in 2013 now... and all those questions have been developed to solutions.
I worked on my DAW prototype for more than two years. It was a technologically advanced machine like no other. If you use technology in the right way... it can allow you to 'see' the future.
That's what supercomputer are for and that's why so many countries spend all that money building supercomputers. But supercomputers are not big and expensive by definition. It is the mindset that it is important.

Inside me, somewhere between technology, emotions, Hans Zimmer's music, Trey Ratcliff's photos and hard reality - methexis was achieved.

Hans Zimmer was tutoring me all along the way with out knowing it.
I was learning from all his success and all the mistakes he was making. (Hans does not afraid to make mistakes.) And there it is once more: Ideas having sex by people who until recently did not know about each-other existence. Every little word he was saying on every little interview, had a meaning for me.

I let you enjoy Hans's Hybrid DAW now.
Hopefully in some point in the foreseeable future nobody will want to work in a DAW like this (Including Hans).

I wish to you a Happy 2013 from the bottom of my heart.
I expect some awesome music from you !
Attached Thumbnails
Building The Most Advanced DAW On The Planet-trey_ratcliff_hans_zimmer_antonis_karalis_crop.jpg  
#72
28th December 2012
Old 28th December 2012
  #72
Lives for gear
 
Joined: Jul 2005
Posts: 3,038

Bassmankr is online now
What specifically are you trying to accomplish? You talk about only having enough money for food and electricity for a period of four years recently and yet in the same post you talk about having a $50k Mac with UAD cards. People barely getting by don't keep $50k in hardware. For under $10k you can assemble a hybrid setup as I describe in my above post where you can have all your instruments available at once for composition. No need for future solutions as it already exists as of today. While evolution of tools is nice there is nothing stopping you right now as a composer with a few dollars. As for Zimmer's setup he needs to network due to employing many composers all using SHARED resources. A single composer would only need some clustered PC's and an analog mixer (even for a 300 track three minute song taking four months to construct). While I suspect things will get better with tech and music software there is nothing stopping anyone from "Get R Done" today.
Antonis2007
Thread Starter
#73
28th December 2012
Old 28th December 2012
  #73
Gear maniac
 
Antonis2007's Avatar
 
Joined: Feb 2007
Location: Mountain View

Thread Starter
Antonis2007 is offline
Quote:
Originally Posted by Bassmankr View Post
What specifically are you trying to accomplish? You talk about only having enough money for food and electricity for a period of four years recently and yet in the same post you talk about having a $50k Mac with UAD cards. People barely getting by don't keep $50k in hardware. For under $10k you can assemble a hybrid setup as I describe in my above post where you can have all your instruments available at once for composition. No need for future solutions as it already exists as of today. While evolution of tools is nice there is nothing stopping you right now as a composer with a few dollars. As for Zimmer's setup he needs to network due to employing many composers all using SHARED resources. A single composer would only need some clustered PC's and an analog mixer (even for a 300 track three minute song taking four months to construct). While I suspect things will get better with tech and music software there is nothing stopping anyone from "Get R Done" today.
You do not have to go to $10K. For under $3K you can setup a small cluster and that's the way thing should be going ! Apart from some 'posts' and some 'random' software developments - we are not going this direction. I want creative freedom, analogue qualities and intuitive tools. That's what I want to accomplish. This will put music in a another era with development as important as the symphonic orchestra. It is time for tools designed with HPC in mind and not with personal computers.

The photo you look at is what Hans use (yes, he share some of that with other composers).
Antonis2007
Thread Starter
#74
28th December 2012
Old 28th December 2012
  #74
Gear maniac
 
Antonis2007's Avatar
 
Joined: Feb 2007
Location: Mountain View

Thread Starter
Antonis2007 is offline
Quote:
Originally Posted by daexpert View Post
I think you're looking for a modular environment such as reaktor, max, or tassman.

I don't think that's what most people want though.
I do not think that's what most people want either!
The last think we need is another layer of complexity.

That reminded me of this 'matrix' of RME sound-cards that keeps surviving on every software update.
#75
28th December 2012
Old 28th December 2012
  #75
Gear nut
 
Joined: Nov 2012
Posts: 96

ArKay99 is offline
Quote:
Originally Posted by daexpert View Post
I think you're looking for a modular environment such as reaktor, max, or tassman.

I don't think that's what most people want though.
No, that's not what I was talking about. I'm talking about a DAW that is more advanced than what is now available. What would be more advanced than being able to 'build' it any way you wish? I simply proposed a way to do it using the technology that currently exists, and by being scaleable it means you could start out with a simple USB interface and expand on that as the budget and aspirations. Even the most advanced front end will need a way to get the music in and out...

Quote:
Originally Posted by Antonis2007 View Post
I do not think that's what most people want either!
The last think we need is another layer of complexity.

That reminded me of this 'matrix' of RME sound-cards that keeps surviving on every software update.
The whole point about abstracting is to avoid any complexity, not add it.

You did propose the concept, "Building The Most Advanced DAW On The Planet"
Antonis2007
Thread Starter
#76
28th December 2012
Old 28th December 2012
  #76
Gear maniac
 
Antonis2007's Avatar
 
Joined: Feb 2007
Location: Mountain View

Thread Starter
Antonis2007 is offline
Quote:
Originally Posted by ArKay99 View Post
No, that's not what I was talking about. I'm talking about a DAW that is more advanced than what is now available. What would be more advanced than being able to 'build' it any way you wish? I simply proposed a way to do it using the technology that currently exists, and by being scaleable it means you could start out with a simple USB interface and expand on that as the budget and aspirations. Even the most advanced front end will need a way to get the music in and out...



The whole point about abstracting is to avoid any complexity, not add it.

You did propose the concept, "Building The Most Advanced DAW On The Planet"
I totally agree with you! That was such a nice post. Thank you for that !
Your input was so... delicately abstracted.

-----------------------------------------------

In general, I think that because of the fact that we are writing posts on the internet - there is a huge gap in the communication. Moreover, good faith is more than welcome when the subject is 'advanced enough' to push all of us to our limits of knowledge, skills and brain juice.

(Also - please forgive my bad English - they are as bad as my Greek - I am a bit dyslectic)
#77
29th December 2012
Old 29th December 2012
  #77
Lives for gear
 
Joined: Jul 2005
Posts: 3,038

Bassmankr is online now
Actually just a single new $1k Sandybridge/Ivybridge computer will be more than enough horsepower for almost all out there making music and recording. Those with the extra bucks can spend more on a mixer and additional PC's which is still very reasonable. Be happy we have so much for so little cost today as it wasn't that far back in time you needed $50k+ to even have a studio. If you want to improve things some more have at it but there are really no road blocks as it stands right now in time.
#78
29th December 2012
Old 29th December 2012
  #78
Lives for gear
 
UnderTow's Avatar
 
Joined: Mar 2006
Location: the Netherlands
Posts: 6,150

Send a message via Skype™ to UnderTow
UnderTow is online now
Hi Antonis,

I didn't read this thread before because I thought it would just be another thread about someone wanting to build the biggest baddest PC for their music DAW. Congratulations on wining the award and I applaud your efforts in trying to improve the tools we all use to make music!

Quote:
Originally Posted by Antonis2007 View Post
In the past we standardized the full size symphonic orchestra (which is a technological masterpiece like no other). It gave us the best sounding acoustic spaces on the planet, it gave us musical instrument technologies.
This created what we now call 'classical music' and it gave us Johann Sebastian Bach and Beethoven.
I believe Beethoven and Johann Sebastian Bach used pen, paper and their minds to create amazing music. Don't miss the forest for the trees.

Quote:
It will have full effects on each stereo track (eq, compressor, delay, reverb, plus one random heavy effect) - all at 192khz full stereo. The effects and the synths would not be the cpu 'friendly' ones... they would be the VSTs that compete 'head to head' to the outboard equivalents. And then I will test it on full discrete surround music project. Because the 'real' orchestra is fully discrete surround for each instrument ! Then 384kHz is phase two of the research project!
I don't mean to offend but I believe that going after sample rates is misguided. I think that anything above 96Khz is a waste of resources. I even think that going over 44.1Khz will have less than 1% or even 0.1% effect on the quality of the sound of your music. And zero effect on the musical quality of the music.

I have heard so many great musicians, performers and composers reminisce about what inspired them to get into music in the first place and so often it is that old crackly AM radio they used to listen to. (Adjust to the era when said person was young). We shouldn't forget that.

IMO this whole idea of throwing more processing power at the problem is not the (main) solution and will come by itself due to the need for more processing power in industries and applications where there is much more money to spend. The real problem isn't lack of processing power (although more processing power is always nice!). I think what we need is better interfaces. By that I mean both 1) Human Computer Interfacing. A way to manipulate sound and music that is intuitive and responsive in a way that allows great human performance and creativity. And 2) better interoperability between systems. There still isn't any widely adopted non-proprietary session interchange or plugin format. So if any way you can harness the power of all those companies to improve how we interface with DAWs and musical instruments and how those DAWs and instruments interface with each other, that would be great!

Of course, there are some great technological developments by some amazing companies like Celemony, iZotope, Zynaptiq etc that make products that were unthinkable even just a decade ago but again, processing power is only enabling things. What really counts is the great minds developing the tools and algorithms to make all this possible. IMO we need more great people working on the tools. More intelligent ideas. Not just more raw processing power. (Within reason. I always want a more powerful PC! :D ).

Unfortunately this has already passed but if it gets repeated, well worth a watch: BBC One - imagine..., Winter 2012, How Music Makes Us Feel

Alistair
__________________
Alistair Johnston - TV & Film Post, Mastering, Sound Design
--
"The first principle is that you must not fool yourself -- and you are the easiest person to fool" -- Richard P. Feynman

"There's a sucker born every minute" -- P.T. Barnum
#79
30th December 2012
Old 30th December 2012
  #79
Gear interested
 
Joined: Aug 2012
Posts: 20

evermindN is offline
Thinking outside the box about ITB music making, it seems to me that the real bottle neck is the interface.
My pet dream is to get my gear hooked up to a box with 24-32 channel ultra low noise amp rig, aka a Quantitative electroencephalogram.
In general complicated interfaces can be very annoying when they "try to read your mind".
I submit that the problem is not that mind reading interfaces ( aka automagic work flows) are unworkable, rather they've been handicapped by obsolete keyboard oriented paradigms.
Antonis2007
Thread Starter
#80
30th December 2012
Old 30th December 2012
  #80
Gear maniac
 
Antonis2007's Avatar
 
Joined: Feb 2007
Location: Mountain View

Thread Starter
Antonis2007 is offline
Quote:
Originally Posted by UnderTow View Post
Hi Antonis,

I didn't read this thread before because I thought it would just be another thread about someone wanting to build the biggest baddest PC for their music DAW. Congratulations on wining the award and I applaud your efforts in trying to improve the tools we all use to make music!

I believe Beethoven and Johann Sebastian Bach used pen, paper and their minds to create amazing music. Don't miss the forest for the trees.

Alistair
Thank you for your wonderful post !

I think that the trees are the 'pen', 'paper' and their 'mind'.
The 'secret' is hidden in their mind.

They used so many things besides their pen, paper and their mind.

I will name a few of them:

They used the best acoustic spaces man ever made - places that have so complex mathematics and 'randomness' that it's 2012 and we do not yet have a algorithm that can represent these complexities.

They used 100 instruments with hundreds of years of technological evolution on them.

They used 100 biological organisms aiming at one direction and operating in real-time (musicians).

They used the conductor - a central point unit that guided them to that direction.

Creative process does not take place when a composer takes a pen and write a melody. It is an evolving experience.
I wish i could know more about it... i am trying really hard to understand where everything is coming from.

----------------------------

To understand how those composers were using their 'mind' and what they were thinking... you have to 'recreate' their environment and 'surroundings'. You can create a fantastic violin melody when for 5 years on a daily basis you 'experience' a violin inside an auditorium. You can compose a breathtaking chord change when for 5 years you experience a strings ensemble inside an auditorium changing chords.

( a quick random thought... for some reason i just remembered Hans Zimmer talking about his 'environment' ... saying that 'my music world is surround')

The environment defines us 99,9% of the time.

The tools and the experiences we have, define the music we create.

There are some rare times that we 'innovate' and we define the environment with our 'innovation' - but even those times... i believe that we are not actually innovate - but it is the environment that creates a 'need' for innovation and we happen to 'support' that need.
------------------------

We just synced a 30 nodes hpc cluster here on HPC Advisory Council. I'm so excited !

I am ready to load 'few' u-he Diva & Zebra at 192kHz and set their 'accuracy' to 'divine'. High sample rate is a true analogue quality ! It's a quality that digital got out of the equation ! Digital was so 'under-powered' at it's first steps that 'it could not exist' if it did not limit itself to the absolute minimum sample rates. Digital gave us a lot of other things! But time passed.... 'time' takes care of a lot of things !

Now the time has come to bring this analogue quality back. That's one reason (not the only one) that i am saying that HPC we will move us from the 'digital' era to the 'virtual' era.

It is time to allow musicians once more to listen to the 'full quality' of their instruments and sounds while they are playing. Allow them to listen to them on marvelous acoustic spaces. and more than that - it is time to give them access to tools that are not limited by the technology of today. HPC is the 'building block' of the future of music and music production.

(ps. Do not underestimate AM radio ! it was 'analogue' and the sound of this radio era had monstrous tube limiters and amplifiers. People who grew up with it..gave us fantastic music and sounds)
Antonis2007
Thread Starter
#81
2nd February 2013
Old 2nd February 2013
  #81
Gear maniac
 
Antonis2007's Avatar
 
Joined: Feb 2007
Location: Mountain View

Thread Starter
Antonis2007 is offline
Hello friends,

These days I am thinking of ways to balance and distribute the power across the nodes of the HPC system.

Allow me to describe what i mean.

SSL 4000 was the first mixing desk that had dynamics on every channel and this feature actually changed the game of music production back then. So with HPC systems it is possible to have full dynamics and EQ on every channel and add time-based effects like reverbs and delays with out making latency an issue.

So if you wanted to specify the channel strip of the 21 century - something that could shape the music of the future, what would you add on it?

Remember, we are in the native wonderland so you could suggest any native plugin you like. I have some ideas and suggestions about reverbs and delays i love but i really like to have your insight on the matter.

Because of the nature of HPC... cpu power is not a limiting factor over here so we love cpu hungry, high quality plugins. We start from 192kHz and are aiming for 384kHz.

If you wanted to create a 300 tracks virtual mixing desk, which would be the plugins that could claim the honor to be included on every channel strip ?
#82
2nd February 2013
Old 2nd February 2013
  #82
Gear addict
 
Joined: Dec 2007
Location: New York City
Posts: 343

jweisbin is offline
Quote:
Originally Posted by Antonis2007 View Post
What would you want from it? .
Being a total geek, I have actually thought about this a lot, and here are some of my ruminations:

being a Mac user, I would want it to at least run on OS X 10.8.x and be 64-bit, but platform independence would of course be nice

totally open source with an SDK so that anyone can contribute to source and plugin design, with a central repository to download branch builds, skins, themes, layouts (see below), plugins, etc.

every single element in the design is user configurable - I mean right down to the location (if fixed), size, color, and type (knob, slider, etc), and mode of operation for tracks, windows, buttons, sliders, menu items, etc. And users can post their skins and layouts to a central site for others to share.

A new way to categorize sound and SFX libraries. IMAO, none to date have made any real progress in usability. For example, look at the names of drum loops in Spectrasonics Stylus - how is one supposed to guess, from the name, what it actually sounds like? A composer is virtually required to audition thousands of loops, sfx, and libraries in his/her arsenal and become familiar with them before becoming really efficient in their use. My suggestion, for drum loops anyway, is to use a syllable-based system like many asian cultures have, like "bish-shapp" means hit the high hat and then close it. One could string together syllables to build up patterns. (BTW I have already patented this so you have been warned :-)) I don't have a solution for pitch/time based material - we need some geniuses to tackle this.

An open source plugin format that would have shells for VST's, AU's, etc (like the "Waveshell"), so contributors could write plugins that would work across all platforms.

In order to add incentive for developers to write these plugins, themes, etc, there could be a central app store type market place where people could try and buy them.

Probably all just pie in the sky, but we can dream.
__________________
Jim Weisbin, C.T.O.
Human Music and Sound Design
#83
5th February 2013
Old 5th February 2013
  #83
Lives for gear
 
Joined: Jul 2005
Posts: 3,038

Bassmankr is online now
"We start from 192kHz and are aiming for 384kHz"

There is no need to use a 192k sampling rate let alone 386k. A higher number does not mean better with sampling rates and for the most part means worse. There is a huge thread here on that subject alone, please educate yourself by reading it or do your own testing to come to the same conclusion. At best SOME converters sound better at 96k. The majority of us record at 44.1k or 48k sample rate here and not because we have to.
Antonis2007
Thread Starter
#84
6th February 2013
Old 6th February 2013
  #84
Gear maniac
 
Antonis2007's Avatar
 
Joined: Feb 2007
Location: Mountain View

Thread Starter
Antonis2007 is offline
Thank you for your reply.

I am aware that some AD/DA OEMs did not manage (yet) to harvest higher sample rate qualities... i talked with them in the past... as i predicted and i am waiting patiently for it... they will introduce a 192khz and a 384khz solution, sooner or later. On the other hand, there are some others like Antelope and DAD that move forward with innovative solutions !

I am also aware that 99% of DAWs users have no real-life option to work on 192kHz in real time on true projects. My prototype DAW down in athens was the most powerful DAW on the planet, until recently (a bit more powerful than Hans Zimmer's and way more technologically advanced). I know that by the reaction from companies that i was asking for Tech-Support when in the past i was trying to do something and i was getting replies like 'We honesty do not know Antonis as non of our customers ever tried this before'. I got that reply to many times the last 4 years.

But now with the research at HPC advisory council we are building systems that have 15 times the power of those DAWs and i can tell you that when you go out of this little box called personal computer, the possibilities are endless. Personal research like the one Hans Zimmer or myslelf do, is good and it can work for the benefit of one person (as a trade secret) but it can not move things forward. That's why someone has to invest the time to make a proper research and suggest solid solutions! I strongly believe that those changes on the mindset of both users and software/hardware OEMS can move music forward.

This project might enjoy the most powerful technology of the world right now...from Mellanox, Intel, AMD, NVidia and 370 companies (http://www.hpcadvisorycouncil.com/co...ers.php)..plus some very forward thinking audio companies which i will talk about soon ... but

the point is not 'processing power' nor the 'gearslutty' nature of HPC and i know it might be the sluttiest project ever made for music production and probably the most ambitious one too.

Thank God there are still some crazy people in this world.

The point is to create solutions that will move music forward - that's our passion and that's our dedication.




Quote:
Originally Posted by Bassmankr View Post
"We start from 192kHz and are aiming for 384kHz"

There is no need to use a 192k sampling rate let alone 386k. A higher number does not mean better with sampling rates and for the most part means worse. There is a huge thread here on that subject alone, please educate yourself by reading it or do your own testing to come to the same conclusion. At best SOME converters sound better at 96k. The majority of us record at 44.1k or 48k sample rate here and not because we have to.
#85
6th February 2013
Old 6th February 2013
  #85
Lives for gear
 
UnderTow's Avatar
 
Joined: Mar 2006
Location: the Netherlands
Posts: 6,150

Send a message via Skype™ to UnderTow
UnderTow is online now
Quote:
Originally Posted by Antonis2007 View Post
Thank you for your reply.

I am aware that some AD/DA OEMs did not manage (yet) to harvest higher sample rate qualities... i talked with them in the past... as i predicted and i am waiting patiently for it... they will introduce a 192khz and a 384khz solution, sooner or later. On the other hand, there are some others like Antelope and DAD that move forward with innovative solutions !
There is absolutely no point whatsoever in recording above 96Khz (and even the worth of that is debatable). Companies like Antelope or DAD either do not understand the most basic fundamentals of how sampling works (unlikely) or are knowingly and consciously misleading their customers. That is nothing to be proud of.

Here are some basics about sampling rates: Digital Audio and Sampling Rates

Quote:
I am also aware that 99% of DAWs users have no real-life option to work on 192kHz in real time on true projects.
It depends what you consider a real project, but much more importantly, there is no point in running a project at 192 Khz in the first place!

Quote:
the point is not 'processing power' nor the 'gearslutty' nature of HPC and i know it might be the sluttiest project ever made for music production and probably the most ambitious one too.
Yet processing power seems to be exactly what you are focusing on. Usability and ergonomics are vastly more important. I'm not sure how the HPC companies can help with that.

Alistair
2
#86
6th February 2013
Old 6th February 2013
  #86
Gear Head
 
Joined: Jul 2012
Posts: 53

james gardener is offline
hey there,
i think that sounds like a very interesting approach!
i like the idea of not beeing forced to seperate eg. composing & mixing. today's only possibilitie to achieve the neccessary cpu power is cluster computers, which are just clumsy and not mobile at all. the idea of modular unlimited cpu power is very tempting.

i always was dreaming of super fast wireless data connections and cloud computing - hey amazon/google, i need power for 40 more nebulas.

my standard channel strip would probably look something like this:
-input/console emulation
-hi pass
-comp/transients
-comp
-fancy plugin
-saturation/tape/distortion
-eq
#87
6th February 2013
Old 6th February 2013
  #87
Lives for gear
 
UnderTow's Avatar
 
Joined: Mar 2006
Location: the Netherlands
Posts: 6,150

Send a message via Skype™ to UnderTow
UnderTow is online now
Quote:
Originally Posted by Antonis2007 View Post
So if you wanted to specify the channel strip of the 21 century - something that could shape the music of the future, what would you add on it?
Only one thing: Enough insert slots to use the plugins of my choice.

Quote:
Because of the nature of HPC... cpu power is not a limiting factor over here so we love cpu hungry, high quality plugins. We start from 192kHz and are aiming for 384kHz.
Ever higher sample rates will not improve the quality of music and is just a huge waste of resources. If you want to make music production available to the masses, this is not the right approach. I feel that you are actually going in the opposite direction of what should be happening in the audio world.

Quote:
If you wanted to create a 300 tracks virtual mixing desk, which would be the plugins that could claim the honor to be included on every channel strip ?
None IMO. Everyone has different needs and preferences. If you insist on including something, I would say a 6 or 8 band EQ/filters, a dynamics section and maybe a saturation tool. That is all.

Btw, I already have a more than 300 track virtual mixing desk. (I've tested my setup and it can run 1000 stereo 96Khz tracks with no issues. This isn't even a new system).

I'm very sorry to say but I don't believe that HPC in itself can really bring much useful to most people in the music or audio world in a direct sense. Indirectly, sure. As technology like Infiniband and massive multi-core support trickles down to consumer products we will see benefits. Running audio applications on a HPC cluster with the latency that that causes? I don't see it happening or being useful outside of experimental stuff. Of course I could be wrong...

Alistair
2
Antonis2007
Thread Starter
#88
6th February 2013
Old 6th February 2013
  #88
Gear maniac
 
Antonis2007's Avatar
 
Joined: Feb 2007
Location: Mountain View

Thread Starter
Antonis2007 is offline
Alister, the latency on my HPC cluster is below 5ms right now... and i am at the first stage of testing.
So it is happening right now ! :-) HPC is changing the game.

and well, 192kHz is barely enough to bring back analogue quality ! but is what we have right now as a common ground.

I this research is so much more than about power and higher sample rate. What i want to achieve is create tools in the virtual world that will move music forward. tools that are not possible to have in the analogue world nor in the personal computer world.

please read the entire thread to get an idea before you post something.

I will soon post an essay about the philosophy behind the research.
Antonis2007
Thread Starter
#89
6th February 2013
Old 6th February 2013
  #89
Gear maniac
 
Antonis2007's Avatar
 
Joined: Feb 2007
Location: Mountain View

Thread Starter
Antonis2007 is offline
now we are talking ! that post is close to the core of the research !
HPC can be small enough - it is more about the philosophy and not the size.

but which plugins? that's what i am trying to investigate right now.
i want this research to be connected with the common ground of tools we all have access right now. some of them are very high quality !

we will build on the available tech and move from that !

so please - read my post #81 Building The Most Advanced DAW On The Planet

and hit me with some insight !


Quote:
Originally Posted by james gardener View Post
hey there,
i think that sounds like a very interesting approach!
i like the idea of not beeing forced to seperate eg. composing & mixing. today's only possibilitie to achieve the neccessary cpu power is cluster computers, which are just clumsy and not mobile at all. the idea of modular unlimited cpu power is very tempting.

i always was dreaming of super fast wireless data connections and cloud computing - hey amazon/google, i need power for 40 more nebulas.

my standard channel strip would probably look something like this:
-input/console emulation
-hi pass
-comp/transients
-comp
-fancy plugin
-saturation/tape/distortion
-eq
#90
6th February 2013
Old 6th February 2013
  #90
Lives for gear
 
UnderTow's Avatar
 
Joined: Mar 2006
Location: the Netherlands
Posts: 6,150

Send a message via Skype™ to UnderTow
UnderTow is online now
Quote:
Originally Posted by Antonis2007 View Post
Alister, the latency on my HPC cluster is below 5ms right now... and i am at the first stage of testing.
So it is happening right now ! :-) HPC is changing the game.
You have a HPC cluster locally (rather than in a cloud)? Do you mind if I ask you a few questions:

What kind of system do you have and how much does such a system cost?
What operating system is it running?
How would it compare to for instance a SuperMicro or Tyan system running Windows HPC Server 2008 with HPC Pack 2012?
Can a standard DAW run on such a system or does it need a proprietary (or at least newly coded) DAW?

This will give us a much better idea of what you are working on.

Quote:
and well, 192kHz is barely enough to bring back analogue quality !
This is absolutely incorrect. Please read through the link I posted above (and especially the further reading like the Lavry paper). If you want to be an audio technology pioneer, you at the very least need to understand how digital audio works. Please do that before making any further steps in what IMHO is the wrong direction.

Quote:
I this research is so much more than about power and higher sample rate. What i want to achieve is create tools in the virtual world that will move music forward. tools that are not possible to have in the analogue world nor in the personal computer world.
I am curious about this but so far you have mentioned sampling rates and 300 tracks virtual mixers. Sampling rates is a misguided direction and 300 track virtual mixers are already readily available. If you are working on something that is truly new, please give us some hints about that. That should be the interesting part.

Quote:
please read the entire thread to get an idea before you post something.
I have Antonis. That is what is worrying me.

Quote:
I will soon post an essay about the philosophy behind the research.
Great!

Alistair
2
Closed Submit Thread to Facebook Facebook  Submit Thread to Twitter Twitter  Submit Thread to LinkedIn LinkedIn  Submit Thread to Google+ Google+ 
 
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Similar Threads
Thread
Thread Starter / Forum
Replies
Dimension Zero / Remote Possibilities in Acoustic Music & Location Recording
11
ISedlacek / Remote Possibilities in Acoustic Music & Location Recording
20
herecomesyourman / Remote Possibilities in Acoustic Music & Location Recording
13
Nevelicious / Remote Possibilities in Acoustic Music & Location Recording
11

Forum Jump

SEO by vBSEO ©2011, Crawlability, Inc.